RAT LIVER MICROSOMES can be shown to carry out. lipid synthesis on added protein. Dependence of microsomal

Transcription

1 Dependence of microsomal lipid synthesis on added protein RUTH TZUR and B. SHAPIRO Department of Biochemistry, The Hebrew University-Hadassah Medical School, Jerusalem, Israel SUMMARY Lipid synthesis by rat liver microsomes from L-a-glycerophosphate-1 -C14 in the presence of palmitic acid, ATP, CoA, and Mgz+ came to a standstill after 30 min while enough cofactors were still present to promote further synthesis. The extent of the reaction depended on the amount of microsomes. Doubling the microsome concentration more than doubled the extent of synthesis. Synthesis of both neutral lipids and phospholipids could be promoted by the addition of inactivated microsomes as well as by serum albumin and by various lipoprotein preparations. A 20- to 40-fold increase of ester yield could be obtained by supplementing low concentrations of microsomes with the proteins. Ovolecithin also had a considerable activating effect. Lipoproteins had an activating effect even in the presence of optimal albumin concentrations, pointing to different modes of action of the two proteins. The hypothesis is discussed that microsomes synthesize lipoproteins and that lipid synthesis is limited by the amount of acceptor protein available. RAT LIVER MICROSOMES can be shown to carry out rapid synthesis of neutral fat and phospholipids, when incubated with a radioactive fatty acid, Coenzyme A, ATP, and Mg2+ (1). Attempts to improve the performance of this system to make it possible to measure net formation of lipid esters were unsuccessful. One limiting factor was the inhibitory effect of one of the substrates-fatty acids-when used in more than millimolar concentration. The second was the relatively short time for which the reaction proceeds at a measurable rate. The present paper deals with this last limitation and with the role of microsomal protein and other added proteins on the extent of synthesis attained. MATERIALS ATP (potassium salt), CoA, sodium a-glycerophosphate, and palmitic acid were commercial preparations. 542 L-a-Glycerophosphate-1 -CI4 was prepared from glyceroll-c14 (Amersham, Bucks, England) and ATP in the presence of glycerokinase (2). Bovine serum albumin (fatty acid poor) was obtained from Pentex Corp., Kankakee, Ill. Cohn s Fractions I + 111, IIIo and IV1 were obtained from the Blood Bank of Magen David, Tel-Aviv, fractionated from human serum according to Method 6 of Cohn (3). Fractions I + I11 and IIIo were dissolved in sodium phosphate (0.05 M) and EDTA (0.005 M) and stored at 4. Fraction IV1 was obtained as lyophilized powder. Fibrinogen and a-globulin were fractions of human plasma obtained from Armour, Chicago, Ill. p-lactoglobulin (bovine) was a crystalline preparation. Purified P-lipoprotein was prepared from human plasma by precipitation with polyvinylpyrrolidone (PVP) accclrding to Burstein (4) or from rat serum brought to density with KBr and centrifuged for 16 hr at 100,000 X g. The floating lipoprotein was dialyzed against 0.9% NaC1. The lipid content in these two types of preparations varied between 5 and 8 peg ester per mg protein. METHODS Native sufiernatant (NS), i.e., the mitochondria-free supernatant fraction of rat liver homogenates in KCl- Tris buffer (9 parts 0.15 M KCl, 1 part 0.5 M Tris-HC1 buffer ph 7.4, M in EDTA), was prepared as described previously (1). Lyophilized Microsomes. The NS was centrifuged at 100,000 X g for 30 min in a Spinco centrifuge at 4. The supernatant fluid was decanted and the microsomal sediment rehomogenized with the KCI-Tris- EDTA solution and centrifuged again for 30 min. The supernatant solution was discarded, the microsomes were suspended in cold double-distilled water, homogenized.

2 and lyophilized. The lyophilized microsomes were stored at -20. Protein was estimated according to Lowry et al. (5). Extraction of Lipids and Fractionation. Chloroformmethanol 2:l (v/v) (12 ml) was added to the reaction mixtures and the lower layer was washed with 8 ml of 0.1 M KC1-methanol-chloroform 94 : 96 : 6 (v/v/v). The chloroform layer contained all the labeled lipids, free of any detectable unchanged a-glycerophosphate. It was evaporated to dryness and the residue was applied in 3 ml of chloroform to a small column of silicic acid (1 g). Neutral lipids were eluted with 50 ml of chloroform and phospholipids with 50 ml of methanol. Delipidation of likoproteins was performed according to the method of Avigan (6) at 4. The lipoprotein solution was adjusted to a specific gravity of with KBr, TI n.- 8 u) 0 E 40 E in order to minimize the denaturation of lipoprotein. Ten volumes of ether were added and the solution was shaken very slowly for 16 hr at 4. The aqueous layer was then dialyzed against 0.9% NaCl in order to remove the KBr and traces of ether. The ether solution was evaporated in the reaction tubes and the lipid dispersed in the reaction mixture by shaking at 37. RESULTS Experiments were performed either with the native supernatant or with lyophilized microsomes. The former had the advantage of simplicity and rapidity of preparation. However, considerable differences in activity were encountered from batch to batch. With lyophilized microsomes, on the other hand, experiments could be repeated with a single preparation for more than two weeks with almost identical results. After 3 weeks activity decreased markedly in these preparations. Here, too, considerable differences were found between various batches, but the activating effects reported were always highly significant. The tables and figures present representative results of at least three experiments. The formation of fatty acid esters by rat liver native supernatant (NS) incubated with ATP, CoA, a-glycerophosphate, Mg2+, and palmitic acid came almost completely to an end after 30 min of incubation. This cessation of activity was not due to lack of reactants, as the minutes FIG. 1. Effect of additional enzyme added at the end of incubation. Native supernatant (NS), 0.1 ml (2.8 mg protein), was added at the beginning of incubation to all flasks, with the indicated additions after 30 min. The incubation mixture consisted of 10 rmoles of ATP, 0.1 mg of CoA, 10 rmoles of MgC12, 20 moles of potassium phosphate buffer ph 7.4, 3.5 rmoles of sodium ~-aglycerophosphate-l-c14 (30,000 cpm), of palmitate, 50 pmoles of KF, and enzyme as indicated in a total volume of 3 ml. Incubation was carried out at 37O under air and incorporation of a-gp (r-a-glycerophosphate) calculated from the percentage incorporation of C1 into lipids. 0 none; X 0.1 ml of NS; ml of NS; A 0.2 ml of NS pmole of palmitate lug microsomal protein FIG. 2. Effect of varying concentrations of lyophilized microsomes on lipid synthesis. Reaction mixtures as in Fig. 1, with lyophilized microsomes added in amounts indicated. Incubation was carried out for 2 hr. TZUR AND SHAPIRO Proteins in Microsomal Lipid Synthesis 543

3 esterification was renewed upon addition of more NS (Fig. 1). With the higher enzyme concentration almost all the palmitic acid was used up and palmitate had to be added for maximal activity. The total esterification TABLE 1 ACTIVATION OF NS ( NATIVE SUPERNATANT ) BY INACTIVATED NS Expt. No. Addition mpmoles a-gp ml of NS ml of preincubated NSt 37 3 As in experiment 2 + fresh 0.2 ml of 195 NS added with cofactors ml of NS 162 ~ ~ Conditions as in Fig. 1 ; NS (3.5 mg of protein in 0.2 ml) was used. Incubation time, 60 min. * a-gp = L-a-glycerophosphate. t 0.2 ml of NS kept at 37 for 30 min. Cofactors and substrates were then added and incubation was continued for 60 min. TABLE 2 EFFECT OF VARIOUS PROTEINS ON LIPID-SYNTHESIZ- ING SYSTEMS IN RAT LIVER MICROSOMES Protein Added mpmoles a-gp None 30 5 mg serum albumin 125 1Omg mg serum Fraction I + I mg ovalbumin 44 10mg 23 5 mg fibrinogen 26 10mg mg 79 5 mg 7-globulin 46 10mg 49 TABLE 3 EFFECT OF VARIOUS LIPOPROTEIN PREPARATIONS ON LIPID SYNTHESIS Enzyme mpmoles Expt. Preparation a-gp No. (NS) Lipoprotein Added ml None mg d < lipoprotein* None (precipi- 98 tated with PVP)* None (Cohn s 206 Fraction 1110) None mg human a-lipoprotein 167 (Cohn s Fraction IVI) mg Cohn s Fraction I Reaction mixture as in Fig. 1. * Lipoprotein preparations are described in Materials JOURNAL OF LIPID RESEARCH VOLUME 5, 1964 obtained after 2 hr of incubation increased with the amount of enzyme and was more than proportional to this amount (Fig. 2). Preincubation of the NS without substrates and cofactors at 37 caused a marked reduction in activity. However, when this inactivated enzyme was supplemented with a fresh portion of NS, the system formed more than the sum of both activities and about as much as would be obtained when starting with twice the amount of unincubated enzyme (Table 1). Small amounts of microsomes (0.9 mg of protein) produced very little lipid. The addition of various proteins increased the activity several times (Table 2). Cohn s Fraction I + I11 and serum albumin were highly active, TABLE 4 Expt. No. PROTEIN EFFECT OF VARIOUS CONCENTRATIONS OF Lipo- AND SERUM ALBUMIN ON LIPID SYNTHESIS Protein Added mg lipoprotein (Fraction I + 111) 6.5 mg 13 mg :: 19.5 mg 26 mg 2 10 mg serum albumin 20 mg 40 mg 3 5 mg 10 mg mpmoles a-gp protein. Incubation carried out for 120 min. TABLE 5 COMBINED EFFECT OF LIPOPROTEIN AND ALBUMIN mpmoles a-gp Albumin f No Albumin Lipoprotein Lipoprotein Addition 10mg 20mg 13mg 26mg 10mg 13 mg protein. Incubation time, 120 min. Quantities of lipoprotein (Cohn s Fraction I + 111) and serum albumin refer to protein content. TABLE 6 ACTIVATION BY ETHER EXTRACT AND BY DELIPIDATED LIPOPROTEIN Expt. No. Addition mpmoles a-gp 1 None Lipoprotein (Cohn s Fraction I + 111) Ether extract of Delipidated lipoprotein of

4 TABLE 7 EFFECT OF LIPID FACTORS ON LIPID SYNTHESIS BY MICROSOMES mpolea a-gp Preparation Addition 1 None 37 Triolein, 3 peq 37 Triolein, 9 peq 37 Sodium cholate, 3 mg 7 Ediol, 2 mg 52 Ediol, 5 mg 52 Ovolecithin, 14 creq None 16 Ovolecithin, 14 peq 171 Ovolecithin, 28 peq 165 protein. Incubation time, 120 min. Ediol = a commercial fat emulsion which contains 50% coconut oil (Schenlabs Pharmaceuticals, Inc., New York). Triolein and lecithin were added as aqueous suspensions prepared by sonication. while y-globulin, fibrinogen and ovalbumin were devoid of activity. An activating effect was obtained also The lack of activity of fibrinogen made it likely that the effect of Fraction I + I11 was based on its other major This could be substantiated by experiments with preparations of purified lipoproteins. All the samples of &lipoproteins, prepared by Cohn s fractionation (Fraction VIIo), by centrifugal flotation, or by precipitation with polyvinylpyrrolidone (PVP), were highly active. Fraction IV1 was also found to activate the synthesis (Table 3). Fraction I + I11 was used in most experiments, as it was the most stable. Optimal activation was obtained with 13 mg of lipoprotein (Fraction I + 111) or with 10 mg of albumin in the reaction mixture (Table 4). The activating effect of serum albumin and lipoprotein was exerted by different mechanisms, as optimal amounts of both activators added together gave additive activating effects (Table 5). Furthermore, while the addition of lipoprotein caused activation at all microsome concentrations tested and corrected the nonlinearity of lipid synthesis found with nonsupplemented microsomes, serum albumin was ef- /ii / pg microsomal protein FIG. 3. Activation by lipoprotein and albumin on varying concentrations of microsomes. Reaction mixtures as in Fig. 1, with lyophilized microsomes added in amounts indicated. 0 microsomes; X microsomes + 10 mg of albumin; A microsomes + 13 mg of Cohn s Fraction I fective only at low enzyme concentrations and was even inhibitory with higher ones (Fig. 3). It seemed of interest to know which part of the lipoprotein was responsible for the activation. 8-Lipoprotein was extracted with cold ether as described in Methods. The ether and the protein portions were then examined for their activity. The ether extract had a significant activating effect on the microsome, but the delipidated lipoprotein proved to be inactive (Table 6). Of the lipids tested, triglycerides, Ediol emulsion and sodium cholate were without effect. Ovolecithin, on the other hand, had a considerable activating effect (Table 7). TABLE 8 ACTNATION EFFECT OF LECITHIN mpmolea a-gp Lecithin Albumin Lipoprotein Lecithin Albumin Lecithin + Lipoprotein Albumin + Lipoprotein Expt. No -- No. Addition 14 w4 10 mg 13 mg 14 peq + 10 mg 14 wq + 13 mg 10 mg + 13 mg a7 Reaction mixture aa in Fig. 1, with 0.9 mg of lyophilized microsome protein. Quantities of lipoprotein (Fraction I + 111) and serum albumin refer to protein content. The quantity of lecithin refers to peq fatty acid ester bonds. TZUR AND SHAPIRO Proteins in Microsomul Lipid Synthesis 545

5 TABLE 9 CENTRIFUGAL SEPARATION OF LABELED LIPIDS - In reaction mixture with 10 mg 13 mg Albumin + Albumin Lipoprotein Lipoprotein Per cent of counts in the supernatant fraction Per cent of counts sedi mented with microsomes protein. The centrifugal separation was pexfonned as indicated in text. The activating effect of lecithin combined with serum albumin or with lipoprotein was tested (Table 8). It was found that lecithin acted additively with lipoprotein, similar to the action of albumin. On the other hand, lecithin was ineffective in increasing the albumin effect. Centrifugal separation (at 100,000 X g for 1 hr) of the reaction mixtures revealed that only in the presence of lipoprotein was a considerable part of the labeled lipids found in the supernatant fraction. In mixtures incubated with albumin almost all the lipid formed precipitated with the microsomes (Table 9). Chromatographic separation of the lipids in the reaction mixtures showed that the synthesis of both neutral FIG. 4. Synthesis of neutral lipids and phospholipids. For details of incubation mixture, see Fig. 1. The protein content of the albumin was 10 mg and of the lipoprotein fraction I was 13 mg. The amount of lecithin was 14 peq. The columns represent two experiments with two preparations: 1-4 and 1-7, with 0.9 mg of lyophilized microsome protein. The blank column represents the phospholipid fraction and the dark column the neutral lipid fraction. 7, microsome without addition; 2, microsome + albumin; 3, microsome + Fraction I + 111; 4, microsome + albumin + Fraction I + 111; 5, microsome + lecithin; 6, microsome + lecithin + albumin; 7, microsome $- lecithin + Fraction I JOURNAL OF LIPID RESEARCH VOLUME 5,1964 lipids and phospholipids was enhanced by the addition of any of the activating factors. The additional synthesis obtained by combining albumin with lipoprotein, or egg lecithin with lipoprotein, was mainly in the neutral lipid fraction (Fig. 4). DISCUSSION The extent of ester formation by rat liver microsomes was shown to be limited by the amount of microsomal protein available. The disproportionately large increase in ester formation when the microsomal protein concentration was increased probably resulted from a combination of factors. For example, dilute microsomal preparations were unstable, and albumin enhanced ru-glycerophosphate esterification in such dilute systems. In contrast, albumin did not have this effect when microsomal concentrations were high. It is possible that albumin stabilized the dilute enzyme preparations just as it has been shown to do for palmitoyl-coa-glycerophosphate transacylase (7). An additional stimulation of glyceride formation was provided by certain lipoproteins. The activating effect of lipoproteins was evident even with higher concentrations of microsomes but was especially impressive with low concentrations. Its mode of action seems to differ from that of albumin, as it was elicited even in the presence of optimal albumin concentrations. The possibility exists that the presence of a protein acceptor is obligatory for the synthesis of lipids, i.e., that the system actually forms lipoproteins rather than free lipids. If the microsomes contain a limited amount of such a protein acceptor, lipid synthesis would stop when this acceptor is saturated. The addition of external lipoproteins or of inactivated microsomes may then serve as additional acceptor or as agents for removing the synthesized lipids from the microsomal protein and allowing it to continue in its synthetic activity. It is at present impossible to decide which of these two mechanisms prevails, although it could be shown that the added lipoprotein incorporates some of the synthesized lipids. When microsomes alone or microsomes with albumin were used, all of the radioactive lipid precipitated with the microsomes ; in the presence of lipoproteins, a considerable part stayed with the lipoprotein in the supernatant fraction. The mode of action of egg lecithin is not yet clear. It promoted the effect of the lipoprotein and not that of the albumin. Because it is a lipid, one would have expected it to replace lipoprotein as extractant of the synthesized lipids. Its activity may be due to the supplementation of the lipid moiety of the lipoprotein or the microsomal lipid acceptor.

6 Ihis investigation was supported in part by PHS Research 2. Tzur, R., E. Tal, and B. Shapiro. Biochini. Biophys. Acta 84: Grant AM from the National Institute of Arthritis and 18, Metabolic Diseases, U.S. Public Health Service. 3. Cohn, E. J., L. E. Strong, W. L. Hughes, Jr., D. J. Mulford, Manusm$t received February 13, 1964; accepted Mav 25, J. N. Ashworth, M. Melin, and H. L. Taylor. J. Am. Chem. Cheni. SOC , Burstein, M. J. Physiol. (Puris) 53: 519, Lowry, 0. H., N. I. Rosebrough, A. L. Farr, and R. J. REFERENCES Randall. J. Biol. Chem. 193: 265, Avigan, J. J. Bid. Ckem. 226: 957, Stein, Y., and B. Shapko. Biochim. Biophys. Acta 24: 197, 7. Brandes, R., J. Olley, and B. Shapiro. Biochm. J. 86: 244; TZUR AND Smmno Proteins in Microsomal Lipid Synthesis 547